摘要
变形作为最直观的监测指标,常用来反映大坝的服役性态变化。为建立更加符合混凝土坝变形的预测模型,实现更高精度的混凝土坝变形预测,针对混凝土坝变形序列呈现不确定性和非线性的特征,将核主成分分析(KPCA)引入最小二乘支持向量机(LSSVM)来约简因子关系,降低预测模型的输入维数和复杂度,同时使用北方苍鹰优化算法(NGO)对最小二乘支持向量机进行参数寻优,构建了基于KPCA-NGO-LSSVM的混凝土坝变形预测模型。工程实例表明,KPCA-NGO-LSSVM模型相比传统多元线性回归(MLR)、LSSVM、KPCA-LSSVM的预测值与实际值的拟合效果更好,预测精度更高,能更有效地预测混凝土坝变形。
As the most intuitive monitoring index,deformation is often used to reflect the change of the service behavior of the dam.In order to establish a prediction model which is more in line with the deformation of concrete dam and realize more accurate prediction of dam deformation,aiming at the uncertain and nonlinear characteristics of deformation sequence of concrete dam,kernel principal component analysis(KPCA)is introduced into least square support vector machine(LSSVM)to reduce the factor relationship and reduce the input dimension and complexity of the prediction model.At the same time,the northern goshawk optimization algorithm(NGO)is used to optimize the parameters of the least square support vector machine,and the concrete dam deformation prediction model based on KPCA-NGO-LSSVM is constructed.The engineering example shows that the fitting effect between the predicted value and the actual value of KPCANGO-LSSVM model is better than that of traditional multiple linear regression(MLR),LSSVM and KPCA-LSSVM,and the prediction accuracy is higher,which can be used to predict the deformation of concrete dam more effectively.
作者
詹明强
陈波
袁志颖
ZHAN Ming-qiang;CHEN Bo;YUAN Zhi-ying(Fujian Shuikou Power Generation Group Co.,Ltd.,Fuzhou 350001,China;College of Water Conservancy and Hydropower Engineering,Hohai University,Nanjing 210098,China;PowerChina Zhongnan Engineering Corporation Limited,Changsha 410014,China)
出处
《水电能源科学》
北大核心
2024年第8期127-131,共5页
Water Resources and Power
基金
国家自然科学基金项目(52079049,52239009)
国家重点实验室基本科研业务费项目(522012272)。
关键词
混凝土坝
核主成分分析
北方苍鹰算法
最小二乘支持向量机
变形预测
concrete dam
kernel principal component analysis
northern goshawk optimization algorithm
least square support vector machine
deformation prediction
作者简介
詹明强(1998-),男,硕士,研究方向为大坝安全监控,E-mail:zhanmingqiang@hhu.edu.cn;通讯作者:陈波(1986-),男,教授、博导,研究方向为大坝安全监控,E-mail:chenbo@hhu.edu.cn。
